Automated image analysis

  • Research type

    Research Study

  • Full title

    Automated image analysis in retinal diseases

  • IRAS ID

    232305

  • Contact name

    Mital Shah

  • Contact email

    mital.shah@nhs.net

  • Sponsor organisation

    Oxford University Hospitals NHS Foundation Trust

  • Duration of Study in the UK

    2 years, 0 months, 1 days

  • Research summary

    Clinical data, such as from retinal imaging, contains a wealth of information. However, its analysis is currently limited by manual techniques. In this study, we will use existing clinical data of patients with retinal diseases to explore a novel methodology for the automated analysis of retinal images. The anonymised clinical data will include information from clinical imaging and functional studies and health records. This novel approach to the analysis of clinical data has already shown some promising results in the detection of diabetic retinopathy from colour retinal photographs and other work in this area is ongoing. Existing algorithms that utilise automated techniques traditionally have very large datasets available with which to develop the algorithms. While this technique is feasible in applications such as data security, access to such large datasets in a real world healthcare setting are impractical. In this study, in addition to utilising the novel methodology of automated image analysis, we will also implement this technique using a much smaller dataset than traditionally used.

  • REC name

    London - West London & GTAC Research Ethics Committee

  • REC reference

    17/LO/1604

  • Date of REC Opinion

    18 Sep 2017

  • REC opinion

    Unfavourable Opinion